Page 262 - Classification Parameter Estimation & State Estimation An Engg Approach Using MATLAB
P. 262

EXERCISES                                                    251

             8. Under which set of assumptions does the EM algorithm degenerate to the K-means
               algorithm? (  )
             9. What would be a simple way to lower the risk of ending up in a poor local minimum
               with the K-means and EM algorithms? (0)
            10. Which parameter(s) control(s) the generalization ability of the self-organizing map
               (for example, its ability to predict the locations of previously unseen samples)? And
               that of the generative topographic mapping? ( )
   257   258   259   260   261   262   263   264   265   266   267